Tau Interactome Mapping Based Identification of Otub1 As Tau Deubiquitinase Involved in Accumulation of Pathological Tau Forms in Vitro and in Vivo

Acta Neuropathol DOI 10.1007/s00401-016-1663-9

ORIGINAL PAPER

Tau interactome mapping based identification of Otub1 as Tau deubiquitinase involved in accumulation of pathological Tau forms in vitro and in vivo

Peng Wang1 · Gerard Joberty3 · Arjan Buist2 · Alexandre Vanoosthuyse1 · Ilie‑Cosmin Stancu1 · Bruno Vasconcelos1 · Nathalie Pierrot1 · Maria Faelth‑Savitski3 · Pascal Kienlen‑Campard1 · Jean‑Noël Octave1 · Marcus Bantscheff3 · Gerard Drewes3 · Diederik Moechars2 · Ilse Dewachter1,4

Received: 2 September 2016 / Revised: 23 December 2016 / Accepted: 23 December 2016 © The Author(s) 2017. This article is published with open access at Springerlink.com

Abstract Dysregulated proteostasis is a key feature of dependent on its catalytically active cysteine, but independ- a variety of neurodegenerative disorders. In Alzheimer’s ent of its noncanonical pathway modulated by its N-termi- disease (AD), progression of symptoms closely corre- nal domain in primary neurons. Otub1 strongly increased lates with spatiotemporal progression of Tau aggregation, AT8-positive Tau and oligomeric Tau forms and increased with “early” oligomeric Tau forms rather than mature Tau-seeded Tau aggregation in primary neurons. Finally, neurofibrillary tangles (NFTs) considered to be pathoge- we demonstrated that expression of Otub1 but not its cata- netic culprits. The ubiquitin–proteasome system (UPS) lytically inactive form induced pathological Tau forms after controls degradation of soluble normal and abnormally 2 months in Tau transgenic mice in vivo, including AT8- folded cytosolic proteins. The UPS is affected in AD and positive Tau and oligomeric Tau forms. Taken together, we is identified by genomewide association study (GWAS) here identified Otub1 as a Tau deubiquitinase in vitro and as a risk pathway for AD. The UPS is determined by bal- in vivo, involved in formation of pathological Tau forms, anced regulation of ubiquitination and deubiquitination. including small soluble oligomeric forms. Otub1 and par- In this work, we performed isobaric tags for relative and ticularly Otub1 inhibitors, currently under development absolute quantitation (iTRAQ)-based Tau interactome map- for cancer therapies, may therefore yield interesting novel ping to gain unbiased insight into Tau pathophysiology and therapeutic avenues for Tauopathies and AD. to identify novel Tau-directed therapeutic targets. Focusing on Tau deubiquitination, we here identify Otub1 as a Tau- Keywords Tau · Tau oligomerization · Tau pathology · deubiquitinating enzyme. Otub1 directly affected Lys48- Alzheimer’s disease · Interactome mapping · Otub1 · linked Tau deubiquitination, impairing Tau degradation, Ubiquitination

Electronic supplementary material The online version of this Introduction article (doi:10.1007/s00401-016-1663-9) contains supplementary material, which is available to authorized users. Alzheimer’s disease (AD) is the most prevalent form of * Ilse Dewachter dementia, being characterized by progressive accumulation [email protected] of amyloid plaques and neurofibrillary tangles, composed of aggregated Aβ and hyperphosphorylated Tau, respec- 1 Alzheimer Dementia Group, Institute of Neuroscience, Catholic University of Louvain, 1200 Brussels, Belgium tively [9, 69]. Over the last decade, Tau has emerged as an important therapeutic target. This is underscored by (i) the 2 Department of Neuroscience, Janssen Research and Development, A Division of Janssen Pharmaceutica NV, close correlation of Tau pathology with the progression of 2340 Beerse, Belgium symptoms and (ii) the existence of a family of Tauopathies 3 Cellzome GmbH, Molecular Discovery Research, which are neurodegenerative disorders characterized by Tau GlaxoSmithKline, Meyerhofstrasse 1, Heidelberg, Germany aggregation spatiotemporally correlating with brain dys- 4 BioMedical Research Institute, Hasselt University, Hasselt, function and associated symptoms, as well as, most impor- Belgium tantly, (iii) the identification of mutations in Tau causally

1 3 Acta Neuropathol linked to Tauopathies, demonstrating a causal executive preference for Lys48-linked polyubiquitin chains over role of Tau in Tauopathies. Although Tau aggregation is Lys11-, Lys29-, or Lys63-linked polyubiquitin chains in closely linked to symptom progression, smaller oligomeric the canonical pathway, dependent on its highly conserved Tau forms rather than mature full-blown neurofibrillary active-site cysteine (C91) [17, 48, 74]. Recently, a nonca- tangles (NFTs) are generally considered to be pathogenetic nonical pathway of Otub1, involved in DNA double-strand culprits in the disease process [6, 11, 12, 15, 20, 33, 37, break response by modulating Lys63 polyubiquitin chain 39, 40, 59, 60, 67, 70]. Accumulation of misfolded and formation, was highlighted [31, 53, 77, 78]. Otub1 has aggregated proteins is a key feature of a variety of neuro- been implicated in a broad range of cellular pathways such degenerative disorders, raising interest in understanding as IL-1β-induced inflammation and stabilization of proteins clearance and degradation mechanisms of these pathoge- including c-IAP1, p53, TRAF3/6, SMAD2/3, and RhoA, netic culprits [7, 22, 28, 35, 42, 43]. The UPS selectively indicating that it is an important regulator in different phys- degrades normal and abnormally folded soluble proteins, iological processes [23, 26, 32, 44, 66], while its role in the which are tagged by ubiquitin for elimination [22], while nervous system remains unknown. the autophagic–lysosomal system (ALS) mainly degrades In this study, we performed iTRAQ-based Tau interac- large protein aggregates or inclusions and organelles. With tome mapping to identify Tau-interacting proteins with smaller soluble Tau forms considered to be toxic culprits, Tau-modifying potential as therapeutic targets. This high- the UPS represents an important target for therapeutic lighted Otub1 as a potential entry point to better understand interventions and is the main focus of this work. the link between the UPS and Tauopathies. We demonstrate The UPS is reported to be downregulated in various neu- here for the first time that Tau is deubiquitinated by Otub1. rodegenerative disorders, with increased proteasome activ- We demonstrate that Otub1 regulates levels of Lys48-linked ity shown to be beneficial in related disease models [15, 28, ubiquitin-conjugated Tau and increases pathological forms 42, 52]. The presence of ubiquitinated Tau in pathological of Tau in vitro and in vivo. lesions of AD provided the first lead implicating the UPS in AD and Tauopathies [49, 50]. In AD patients’ brain, paired helical filaments (PHFs) were found to be associated Materials and methods with impaired proteasome activity in a brain-region specific way [15, 28, 34, 35]. Furthermore, a ubiquitin mutant Animals 1 called UBB+ , with a 19-AA extension, has been identified in neurons from AD patients and suppresses UPS func- TauP301S (PS19) mice [81] were bred and used in our labo- 1 tion [38, 45, 71]. UBB+ -induced proteasome dysfunction ratory as reported previously [64, 65, 72]. The mice bred in resulted in Tau pathology and related neuronal dysfunction our laboratory develop Tau pathology and a neurodegenerative 1 in transgenic mice expressing UBB+ [29]. Accumulating phenotype starting at around ~10–11 months. In this study, evidence indicates that the UPS is implicated in clearance stereotactic injections were performed at P0 and age-matched of different Tau forms and its dysfunction is closely related littermates were used for analysis 2 months postinjection. At to Tau aggregation and accumulation [1, 13–16, 25, 46, 57, the age of 2 months, no Tau pathology or AT8-positive stain- 68]. Conversely, Tau accumulation was recently shown to ing is detected in TauP301S mice in our colony. All experi- impair proteasomal degradation, suggesting a vicious cir- ments were performed in compliance with protocols approved cle [52]. Degradation of target proteins through the UPS is by the UCLouvain Ethical Committee for Animal Welfare. regulated by ubiquitination, with differently linked (Lys6, Lys11, Lys48, and Lys63) polyubiquitinated Tau identified Reagents and antibodies in AD brains and models [13, 49–51, 56]. Different Tau forms can be ubiquitinated by E3 ligases CHIP [57, 63], MG132 and cycloheximide were purchased from Sigma. TRAF6 [1], and MARCH7 [19], with particularly CHIP/ Phosphatase inhibitor (PhosSTOP™) and protease inhibi- Hsp70-dependent polyubiquitination characterized in detail tor cocktails (cOmplete™, Mini, EDTA-free) were obtained in vitro and in vivo. Ubiquitination is highly dynamic and from Roche. Primary antibodies used in this study included reversible by the action of deubiquitinating enzymes known antibodies directed against human Tau (Dako), Otub1, Ubiq- as deubiquitinases (DUBs), hence being determined by bal- uitin-Lys48-specific, Ubiquitin-Lys63-specific (Abcam), anced regulation between ubiquitination and deubiquitina- Phospho-Tau (Ser202/Thr205) AT8, Phospho-Tau (Thr231) tion. While Tau ubiquitination has been studied intensively, AT180, Phospho-Tau (Thr212/Ser214) AT100 (Thermo its deubiquitination remains less well explored. Fisher Scientific), and Oligomeric-Tau T22 (Merck), applied Deubiquitinases are attracting increasing interest as in combination with appropriate horseradish peroxidase therapeutic targets. Otub1 is an ovarian-tumor domain (HRP) or Alexa Fluor® (488/568/647) coupled secondary cysteine protease deubiquitinating enzyme [2] with strong antibodies.

1 3 Acta Neuropathol

Cell culture and transfection For Western blot analysis, cells were washed twice with PBS and extracted for 30 min at 4 °C with Triton lysis Human kidney-derived QBI-293 were originally obtained buffer (1% Triton X-100, 50 mM Tris, 150 mM NaCl, pH from QBiogene (Carlsbad, CA, USA) and HEK293 cells from 7.6) containing protease and phosphatase inhibitors, and American Type Culture Collection (ATCC), and were cultured centrifuged at 12,000 g for 15 min at 4 °C to remove × according to the manufacturer’s protocol. All cells were main- insoluble material. Protein content was determined by BCA tained at 37 °C in humidified atmosphere with 5% CO2. Cells Protein Assay kit (Thermo Fisher Scientific, Waltham, MA, were transfected with plasmids expressing Otub1, USP9x, USA). Samples (10 μg) were separated using precast 8% USP5 or empty vector using FuGENE® 6 (Roche) or short Tris–glycine gels or 4–12% Bis–Tris gels (MOPS running; interfering RNAs (siRNAs) using Lipofectamine 2000 (Invit- Invitrogen) and transferred to polyvinylidene difluoride rogen) according to the manufacturer’s protocol. For analysis membranes. Immunoblotting was performed using the of the effects on Tau or on Tau-seeded Tau aggregation, trans- indicated primary antibodies with corresponding second- fection was performed 24 h before starting the assay. ary antibodies, and developed using ECL kit (PerkinElmer, Waltham, MA, USA). In vivo gene delivery Tau aggregation assay Adeno-associated viral (AAV) vectors expressing wild- type Otub1 fused with a green fluorescent protein (GFP) Tau PFFs (synthetic preformed fibrils), referred to as “Tau tag (AAV–Otub1) or expressing GFP (AAV–GFP) driven seeds,” were generated as described previously [24, 65, by neuron-specific Syn promoter were generated using a 72]. Truncated human Tau fragments bearing a proaggrega- previously described plasmid (Addgene, plasmid #26976). tion mutant (Tau-P301L) containing the four-repeat domain Mutant Otub1 C91A and N-terminal truncation (N-T) were [K18; Q244-E372 (4RTau)], N-terminally Myc-tagged were generated by molecular cloning starting from the WT- produced in Escherichia coli (TEBU-Bio). Tau-PFFs were Otub1 construct. AAV viruses were produced in HEK-AAV obtained by incubation of Tau fragments (66 μM) at 37 °C cells using AAV-DJ8 kit (Cell Biolabs, Inc.), and subse- for 5 days in presence of heparin (133 μM) in 100 mM quently purified and concentrated by iodixanol-based ultra- ammonium acetate buffer (pH 7.0), spun down (100,000 g, × centrifugation. AAV titer was tested using a QuickTiter™ 1 h, 4 °C), resuspended to 333 μM, and sonicated before use. AAV Quantitation Kit (Cell Biolabs, Inc). For P0 injection, In vitro Tau aggregation assay in HEK293 cells: PFF- each mouse was injected into the lateral ventricles of both induced Tau aggregation in vitro was performed essentially cerebral hemispheres with 4.2 109 total viral particles as described previously [24, 65, 72]. Sonicated Tau seeds × per side, TauP301S transgenic mice were euthanized, and were diluted in 100 mM ammonium acetate buffer (pH brains were dissected as described previously following 7.0) to 10 µM solution and sonicated with 8 pulses/30% transcardial flushing and analyzed at 2 months postinjec- amplitude and added to the cells using BioPORTER® tion [64, 65, 72]. (AMS Biotechnology, Milton, UK) according to previously described protocol. To detect Tau aggregation, cells Immunoprecipitation and Western blot analysis were fixed with 4% paraformaldehyde (PFA) containing 2% sucrose and 1% Triton X-100 for 15 min to remove To detect Tau ubiquitination, primary neurons were infected soluble proteins. After washing with PBS, aggregated Tau- with AAV–Otub1 (wild type; catalytically dead mutant GFP was analyzed microscopically. To detect the effect of C91A; N-terminal truncation) or AAV–GFP, five days after Otub1, USP5, and USP9x overexpression or knockdown infection, cells were washed with phosphate-buffered saline on Tau aggregation, stably Tau-expressing QBI-293 cells (PBS) and lysed in TGN lysis buffer (50 mM Tris HCl, pH were transfected with plasmid and empty vector or siRNA 7.5, 200 mM NaCl, 50 mM sodium β-glycerophosphate, in 24-well plates. At 24 h after transfection, the growth 1% Tween 20, 0.2% NP40) containing phosphatase inhibi- medium was replaced with OptiMEM, and sonicated Tau tor (PhosSTOP™; Roche) and protease inhibitor cocktails seeds (10 μM) were added to BioPORTER® single-use (cOmplete™, Mini, EDTA-free; Roche) at 4 °C for 30 min tubes and added to cells. Three days after seeding, Tau- on a wheel rotor, and spun at 12,000 g for 15 min. Fol- GFP aggregation was measured by microscopy. × lowing preclearing without antibodies at 4 °C for 1 h, the In vitro Tau aggregation assay in primary neurons was supernatants were incubated with specific antibodies for performed as described previously [65]. Tau seeds (10 nM) 1 h at 4 °C, followed by incubation with protein A-Sepha- were added at DIV3 and DIV6 to primary cortical neu- rose beads at 4 °C for 45 min. Following stringent washing ronal cultures (PNC) from P0 heterozygous TauP301S with TGN buffer and PBS, immunoprecipitated proteins pups. To detect the effect of Otub1 on Tau aggregation in were analyzed by immunoblotting. primary neurons, AAV–Otub1 and AAV–GFP infections

1 3 Acta Neuropathol were carried out at DIV3, Tau seeds (10 nM) were added Tau. Retrieval of several well-characterized Tau-interacting at DIV6 and DIV9, and cells were fixed at DIV12 with proteins validated the experimental approach (Fig. 1b). 4% PFA containing 2% sucrose and 1% Triton X-100 for These included several tubulin isoforms [41, 75, 79], dif- 15 min to remove soluble proteins. After washing with ferent components of the dynactin complex [47], enzymes PBS, aggregated Tau-GFP was detected under microscope. modifying Tau phosphorylation, and members of the heat shock protein family, i.e., Hsc70 (Hspa8), Hsp70 (Hspa4), Immunofluorescence microscopy assay and Hsp110 (Hsp110) [18, 30, 57]. In addition, we identified many new putative Tau interactors, based on their At indicated times after infection, cells were washed by significant enrichment in anti-Tau immunoprecipitation phosphate-buffered saline (PBS) for three times and fixed compared with control immunoprecipitations (Tables S1, with 4% paraformaldehyde in PBS for 15 min at room tem- S2). The full Tau interactome is presented in Supplemen- perature (RT). For stringent extraction to detect aggregated tary Tables S1–S4. Gene ontology and STRING analysis Tau, 1% Triton X-100 (4% PFA, 2% sucrose) was used. were performed (Fig. S1; Table S3), and candidate interac- Cells were briefly permeabilized in PBS/0.2% Triton X-100, tors grouped according to their function (Fig. 1c; Table S4) then blocked in blocking solution (PBS containing 10% fetal to yield new insights into the pathophysiological role of calf serum and 0.1% Triton X-100). Primary and secondary Tau. Within this project, we focused on proteins involved antibody incubations were performed in blocking solution in regulation of Tau ubiquitination and its degradation by overnight at 4 °C or 1 h at RT using the indicated antibodies the ubiquitin proteasome system, generally considered to and goat anti-mouse IgG1 or goat anti-rabbit IgG1 secondary be involved in degradation of small soluble proteins. We antibody coupled to Alexa® 488, Alexa® 568 or Alexa® 647. focused particularly on Tau deubiquitination and identified Cells were visualized using a digital inverted fluorescence three major candidate deubiquitinating enzymes, including microscope (EVOS-xl auto microscope). Otub1, USP5, and USP9x, with Otub1 as the strongest Tau interactor according to statistical analysis. Tau interactome mapping Otub1 increases Tau stability and increases Tau Total mouse brain homogenates were used for coimmunopre- aggregation in a well‑characterized cellular Tau cipitation in combination with three different polyclonal anti- aggregation model Tau antibodies and control antibodies that were cross-linked to Sepharose beads. Mass spectrometry procedures were To analyze whether Tau-interacting proteins, identified by performed essentially as previously reported [4]. A detailed the Tau interactome mapping, exerted modifying effects description of the Tau interactome mapping procedure is pro- on Tau aggregation, we used a well-validated cellular vided in the Supplemental Experimental Procedures. Tau seeding model [24, 65, 72]. We analyzed the effect of the three identified candidate DUBs on Tau aggregation in vitro. In this assay, QBI cells stably expressing the Results longest human Tau isoform (2N4R), with the aggregation- prone P301L mutation and GFP tagged, were used in Tau interactome mapping identifies Tau‑interacting combination with preformed preaggregated synthetic Tau proteins in mouse brain, including Otub1 as a potential seeds consisting of Tau fragments harboring P301L mutant Tau modifier (myc-K18/P301L) to induce monomeric Tau aggregation. In this assay, addition of Tau seeds induces robust aggrega- To identify Tau-interacting proteins with Tau-modifying tion of monomeric human Tau, as assessed by cytological potential, we performed Tau interactome mapping of analysis following permeabilization with 1% Triton X-100 endogenous Tau in mouse brain (Fig. 1) using a proteomic to remove soluble forms of Tau while retaining aggre- approach [3, 5]. Hereto, iTRAQ-based quantitative mass gated hyperphosphorylated Tau (Fig. 2a), as previously spectrometry was used in combination with differential demonstrated and confirmed biochemically (Fig. S2) [65, coimmunoprecipitation of endogenous Tau from mouse 72]. Expression of the candidate deubiquitinases Otub1, brain. The use of isobaric isotope labels allows side-by- USP5, and USP9x in this assay demonstrated significantly side quantitative comparison of up to four different samples increased Tau aggregation following expression of Otub1 [76] (Fig. 1a). Four Tau immunoprecipitation experiments (Fig. 2a; Fig. S3), not observed following expression of were performed using three different anti-Tau antibod- USP5 or USP9x (Fig. S3). Next, we knocked down Otub1 ies and two different control non-Tau-capturing antibod- in the Tau seeding assay, using two siRNAs of Otub1 with ies for differential analysis (Fig. 1a). Statistical tools were different knockdown efficiency. Endogenous Otub1 deple- applied to score proteins for selective coprecipitation with tion significantly inhibited Tau aggregation, with the level

1 3 Acta Neuropathol

Fig. 1 iTRAQ mass spectrometry-based Tau interactome mapping in mouse brain identifies well-characterized and novel Tau-interacting proteins. a Four immunoprecipitation (IP) experiments were performed using three anti-Tau and two control antibodies as indicated (left panel iTRAQ experimental layout; right panel detail of the four experiments performed). b Graphical display of anti-Tau immunoprecipitation (IP) data. Results of anti-Tau hTau24 (X-axis) IPs are compared with results of anti-Tau hTau21 IPs (Y-axis). Results are shown as log2 of ratio between quantification of protein immunoprecipitated with anti-Tau and control antibodies. Only proteins specifically precipitated with anti- Tau antibodies are displayed [known Tau interactors (green dots); Otub1 (purple dot)]. Significantly enriched proteins with only one of the two antibodies are artificially drawn below the axes. Statistical tools were applied to identify a list of Tau-interacting proteins from the differential co-IP analysis in the four experiments. c Fol- lowing gene ontology analysis, candidate Tau interactors were grouped according to their function; a weighted (font size) presentation of the interactors (gene nomenclature) is displayed. The complete list of Tau-interacting proteins is provided in Tables S1–S4, and STRING analysis in Fig. S1

of inhibition correlating with the level of knockdown effi- Since Otub1 is a deubiquitinase, known to regulate the ciency, while use of a control siRNA did not affect levels stability of many crucial proteins in different cellular pro- of Otub1 nor aggregation efficiency (Fig. 2b). Combined, cesses [23, 26, 32, 44, 66], we assessed whether Otub1 these data indicate that Otub1 affects Tau aggregation in a affected Tau stability by measuring Tau half-life following well-characterized Tau-seeded Tau aggregation model. addition of the protein synthesis inhibitor cycloheximide

1 3 Acta Neuropathol

◂Fig. 2 Otub1 increases Tau stability and increases Tau aggregation Tau seeds allows bypass of the lag phase of Tau aggrega- in a well-characterized cellular Tau aggregation model. a QBI-293 tion, resulting in recruitment of soluble Tau into insoluble cells, stably expressing GFP-tagged TauP301L, were seeded with preaggregated synthetic Tau seeds to induce monomeric Tau aggre- Tau aggregates, reflected in a robust detergent-resistant AT8 gation. Tau aggregation was assessed by stringent extraction using (phosphorylated Tau) signal throughout soma and neurites 1% Triton X-100 to remove soluble proteins. Tau aggregation follow- (Fig. 4a, Fig. S2a). We used this well-characterized assay ing expression of Otub1 and control vector in QBI-293 cells reveals to analyze a potential effect of Otub1 on Tau aggregation significant increase of Tau aggregation following Otub1 expression, demonstrated by quantitative analysis (n > 36 fields per condition; in TauP301S cortical primary neurons using AAV-driven scale bar 200 μm; **p value <0.01). Otub1 overexpression was bio- expression of Otub1 or GFP for control. Otub1 expres- chemically validated using anti-Otub1 antibody. b Endogenous Otub1 sion strikingly enhanced detergent-resistant AT8-positive depletion using two different siRNAs of Otub1 significantly inhibited Tau accumulation compared with GFP-infected neurons Tau aggregation, correlating with the level of knockdown efficiency, as demonstrated by quantitative analysis (n > 36 fields per condi- (Fig. 4b), indicating that Otub1 increased Tau-seeded Tau tion; scale bar 200 μm; **p value <0.01; ***p value <0.001). Otub1 aggregation in primary neurons, corroborating the results knockdown efficiency was quantified by Western blotting using anti- obtained in a nonneuronal cell line. Otub1 antibody. c Endogenous Otub1 depletion using two different To measure the effect of Otub1 expression on oligomeric siRNAs of Otub1 significantly increased Tau turnover, measured following treatment with cycloheximide (CHX, 20 μg/ml) for indicated Tau forms, we performed Western blotting analysis in non- times to inhibit protein synthesis and analyzed by immunoblotting. reducing conditions, following AAV-driven expression of Actin served as loading control within each group (*p value <0.05; Otub1 and GFP, in absence of Tau seeding. This revealed **p value <0.01) a significant increase of oligomeric Tau forms in AAV– Otub1-infected neurons compared with AAV–GFP-infected (CHX). Knockdown of Otub1 increased Tau turnover, cor- neurons (Fig. 4c). In addition, we found robust induction relating with the decrease in expression of Otub1 obtained of a (homo- or heterotypic) Tau complex (Tauc), which by the different siRNAs (Fig. 2c). Taken together, our data was strongly detected in AAV–Otub1-infected neurons but indicate that Otub1, a deubiquitinating enzyme, is a novel absent in AAV–GFP-infected neurons, indicating clearcut positive regulator of Tau aggregation and Tau stability Otub1-induced modulation of Tau in neurons. The induc- in vitro, in a nonneuronal cell line. tion of oligomeric Tau forms was further confirmed using dot-blot analysis with T22 antibody (Fig. 4d), revealing a Otub1 increases total Tau concentration and Tau significant increase in oligomeric Tau following expression phosphorylation in primary neurons of Otub1. Taken together, our data demonstrate that Otub1, iden- We next analyzed the modulatory effect of Otub1 on Tau in tified as a Tau-interacting protein in the Tau interactome primary neurons, as relevant cell type for the study of Tauopa- mapping, modulates Tau, by increasing Tau-seeded Tau thies and AD. Hereto, primary cortical neurons, derived from aggregation, by formation of a Otub1-induced Tau complex TauP301S transgenic pups (PS19) [65, 81], were infected Tauc, and by increasing the concentration of oligomeric Tau with AAV–Otub1–GFP or AAV–GFP driving neuron-specific forms in primary neurons. expression. Immunocytological analysis with AT8 antibody against phosphorylated Tau (Ser202/Thr205) demonstrated Otub1 but not Otub1‑C91A impairs Tau degradation that Otub1 expression drastically increased phospho-Tau by removing Lys48‑linked polyubiquitin chains (p-Tau Ser202/Thr205) in primary neurons compared with from Tau in primary neurons GFP expression (Fig. 3a). Biochemical analysis confirmed a sharp increase of phosphorylated Tau at Ser202/Thr205 in Previous studies have shown that Otub1 has dual functions Otub1-infected primary neurons (Fig. 3b). Otub1 also signifi- in regulating both ubiquitin assembly and disassembly [77, cantly increased total Tau protein level compared with GFP 78]. Otub1, as an isopeptidase, can remove Lys48-linked control (Fig. 3b). This increase was, however, less robust than polyubiquitin chains from its substrates [17, 48, 74], but the increase in AT8-positive Tau, as reflected by the ratio of can also serve as an E2 enzyme inhibitor to impede Lys63- AT8/total Tau. Taken together, we conclude that overexpressed specific polyubiquitin chain formation in a noncanonical Otub1 in primary neurons expressing TauP301S increased manner [61]. Since Otub1 interacts with Tau and promotes total Tau and induced a robust increase in AT8-positive Tau. phosphorylated and aggregated Tau levels in primary neurons, we hypothesized that Otub1 could act as a Tau deu- Otub1 enhances Tau oligomerization and Tau‑seeded biquitinase, interfering with pathological Tau degradation Tau aggregation in primary neurons and hence Tau aggregation. To test this assumption experi- mentally, we analyzed the ubiquitination status of Tau by We next analyzed the effect of Otub1 in a validated Tau immunoprecipitation. Following AAV-driven expression of aggregation assay in primary neurons [65, 81]. Addition of Otub1 or GFP in primary neurons, Tau proteins were pulled

1 3 Acta Neuropathol

Fig. 3 Otub1 increases total Tau concentration and Tau phosphoryla- for each infection, three repetitions are analyzed). b Western blot tion in primary neurons. a Immunofluorescence staining of primary analysis of AAV-infected primary TauP301S neuron lysates using TauP301S neurons infected with AAV–Otub1 or AAV–GFP using AT8 and total Tau antibodies. Actin was used as loading control on AT8 recognizing Tau phosphorylated at the pathologically relevant each blot. Ratio of AT8/Tau indicates the preferential effect of Otub1 epitope Ser202/Thr205. Higher magnifications of AT8-stained neu- on Phospho-Tau (AT8) (*p value <0.05; **p value <0.01; ***p value rons of Otub1- and GFP-expressing neurons are presented. Quantifi- <0.001; n 3 different infections; for each infection, three repeti- cation of staining intensities are presented as mean standard error tions were analyzed)= of the mean (SEM) (***p value <0.001; n 3 different± infections; =

1 3 Acta Neuropathol

Fig. 4 Otub1 enhances Tau oligomerization and Tau-seeded Tau analyzed; scale bar 200 μm; ***p value <0.001). c Western blotting aggregation in primary neurons. a Tau aggregation assay in TauP301S analysis of primary neuron extracts in nonreducing conditions, fol- primary neurons. Addition of Tau seeds resulted in phosphorylated lowing AAV-driven expression of Otub1 and GFP, reveals increased Tau (Ser202/Thr205) staining in soma and neurites, resisting strin- monomeric Tau and increased oligomeric forms of Tau (detected as gent extraction with 1% Triton X-100 (n > 36 fields per condition; high-MW smears) in Otub1-overexpressing neurons (n 3 different scale bar 200 μm; ***p value <0.001). b Primary TauP301S neurons experiments; for each infection, three repetitions were =analyzed; *p were infected with AAV–Otub1 or AAV–GFP three days after cultur- value <0.05; **p value <0.01). Induction of a Tau complex (homo- ing, and seeded 1 and 4 days postinfection, respectively. At 7 days or heteromeric)—denoted Tauc—is clearly noticed following AAV– after transduction, neurons were fixed and stringently extracted Otub1 but not AAV–GFP infection. d Dot-blot analysis of primary for analysis of aggregated Tau. Immunofluorescent staining using neuron lysates using oligomeric Tau-specific antibody T22 was per- AT8 (Ser202/Thr205) after extraction is shown, demonstrating Tau formed and quantitatively analyzed following normalization to total aggregation. Quantification is presented as mean SEM (n 3 dif- Tau (dot-blot analysis) (mean SEM; n 8 infection for each ferent experiments; for each infection, 36 fields± per condition= were group; *p value <0.05) ± =

1 3 Acta Neuropathol

Fig. 5 Otub1 but not Otub1-C91A increases Tau stability by remov- chain formation was detected using Ub-Lys48 antibody. c Primary ing Lys48-linked polyubiquitin chains from Tau in primary neurons. TauP301S neurons were infected with AAV–Otub1 compared with a Primary neurons from TauP301S mice were infected with AAV– AAV–GFP. Five days later, neurons were treated with 10 µM protein Otub1 or AAV–GFP. At 7 days after infection, Tau proteins were synthesis inhibitor cycloheximide (CHX) for different durations up immunoprecipitated with Tau antibody and analyzed by linkage- to 36 h as indicated. Similar comparison was performed for primary specific ubiquitin (Ub-Lys48, Ub-Lys63) immunoblotting. A repre- neurons from TauP301S mice infected with AAV–Otub1-WT com- sentative blot is shown; quantification of three different experiments pared with AAV–Otub1-C91A. The rate of Tau turnover was quan- is presented (***p value <0.001). b Otub1-WT or catalytically dead tified using Western blot analysis. Actin served as loading control mutant (C91A), N-terminal truncated mutant (N-T), and GFP control within each group. Quantification was performed on three independ- were expressed in primary TauP301S neurons. Immunoprecipitation ent experiments (**p value <0.01; ***p value <0.001) was performed using Tau antibody, and Lys48-linked polyubiquitin

1 3 Acta Neuropathol down, and Lys48- and Lys63-linked polyubiquitination of of Otub1, which is crucial for its noncanonical role, is not Tau was measured by linkage-specific ubiquitin antibod- involved in regulation of Tau ubiquitination. ies. To enrich ubiquitinated proteins, proteasome inhibitor To confirm the effect of Lys48-linked polyubiquitination was added to the medium 6 h before harvesting. Otub1 dis- change on Tau degradation, TauP301S neurons were infected played strong preference for disassembling Lys48-linked with AAV–Otub1–GFP or AAV–GFP, and subsequently treated polyubiquitin chains, while leaving Lys63-linked polyubiq- with protein synthesis inhibitor cycloheximide (CHX) for dif- uitin chains unaffected (Fig. 5a). Analysis of a catalytically ferent durations up to 36 h to quantify the rate of Tau turnover dead mutant with a C91A mutation revealed no effect on using Western blot. We found that Tau degradation was signifi- Lys48-linked polyubiquitination of Tau. Conversely, AAV- cantly impaired in primary neurons infected with Otub1 WT, driven expression of N-terminally truncated Otub1 sig- but not with catalytically dead mutant C91A, compared with nificantly decreased Lys48-linked polyubiquitin chains of GFP control. These results indicate that Otub1 regulates Tau Tau (Fig. 5b), demonstrating that the N-terminal domain degradation, dependent on its catalytic activity (Fig. 5c).

Fig. 6 The effect of Otub1 on Tau phosphorylation is dependent on Otub1 catalytic activity. a Immunofluorescence staining of TauP301S primary neurons infected with AAV–Otub1 wild type (WT), catalytically dead mutant (C91A) or N-terminal truncation (N-T) using phospho- Tau antibody AT8 (Ser202/ Thr205). Quantification of staining intensities is presented as mean SEM (***p value <0.01; n ± 3 different infections; for= each infection, three repetitions are analyzed). b Western blot analysis of AAV- infected primary neurons lysates using AT8 antibody. Actin was used as loading control (**p value <0.01; n 3 different infections; for each= infection, three repetitions were analyzed)

1 3 Acta Neuropathol

We further analyzed whether catalytically dead Otub1- Fig. 7 Otub1 increases concentrations of phosphorylated Tau and ▸ C91A could affect Tau phosphorylation in primary neurons Tau oligomers in vivo in Tau transgenic mice. a Immunohistochemi- cal staining of cortex of TauP301S transgenic mice injected with expressing TauP301S. AAV-driven expression of Otub1- AAV–Otub1 or AAV–GFP at P0 and analyzed at 2 months. Repre- C91A did not increase AT8-positive Tau, in contrast to sentative images of AAV-driven expression of Otub1 and GFP are expression of wild-type Otub1 and the N-terminally trun- presented (green). Representative AT8 stainings are presented (red). cated form of Otub1. This was demonstrated by immuno- Quantitative analysis is presented as mean SEM (n 9 mice for each group; 3–5 sections were examined for± each mouse;= 9–12 fields fluorescence staining (Fig. 6a) and confirmed by biochemi- were quantified for each section; scale bar 400 μm; ***p value cal analysis (Fig. 6b). <0.001). b Western blotting analysis of phosphorylated Tau and total Taken together, our data prove that Otub1 can remove Tau in cortex homogenates of AAV–Otub1- and AAV–GFP-injected Lys48-linked polyubiquitin chains from Tau in a catalytic mice, using AT8 (Ser202/Thr205) and anti-Tau antibody, is presented. Quantitation was performed on n 9 mice for each group; the AT8/ activity-dependent manner, implicating the canonical path- Tau ratio indicates the preferential= effect of Otub1 on Tau phospho- way. Consequently, Otub1-dependent deubiquitination rylation (*p value <0.05). c Western blotting analysis of mice cor- inhibits Tau degradation through the proteasome pathway tex homogenates in nonreducing condition reveals clear induction and prolongs its half-life, leading to accumulation of patho- of oligomeric forms of Tau (detected as high-MW smears), as well as monomeric Tau protein level increase in TauP301S mice injected logical forms of Tau. with AAV–Otub1 (n 9 mice for each group; *p value <0.05; **p value <0.01). Induction= of a Tau complex (homo- or heteromeric) Otub1 promotes Tau phosphorylation of ~120 kDa—denoted Tauc—is clearly detected following AAV– and oligomerization in vivo in Tau transgenic mice Otub1 but not AAV–GFP infection. d Dot-blot analysis of mice cortex homogenates using oligomeric Tau-specific antibody T22 is presented. Quantitative analysis was performed following normalization To investigate the effect of Otub1 on Tau in vivo, AAV–Otub1 to total Tau (dot-blot analysis) (n 6 mice for each group; *p value and AAV–GFP were bilaterally intraventricularly injected <0.05) = into TauP301S transgenic pups (P0), resulting in widespread expression of Otub1 and GFP in cortical neurons at 2 months of age (Fig. 7a). Immunostaining with AT8 antibody revealed in vivo, we performed P0 injections of AAV–Otub1-C91A abundant AT8-positive neurons in AAV–Otub1-injected mice, or AAV–Otub1-N-terminal truncated mutant, and AAV– absent in AAV–GFP-infected mice (Fig. 7a), at 2 months Otub1-WT in TauP301S mice. Immunostaining and bio- postinjection. Biochemical analysis further confirmed chemical analysis, 2 months postinjection, unambiguously increased AT8-positive Tau by Otub1 expression (Fig. 7b), revealed that the catalytically dead mutant C91A did not with AT8 phosphorylated Tau more robustly increased than affect AT8-stained Tau in vivo, while N-terminal truncated total Tau, as reflected by the AT8/total Tau ratio. Cortex mutant of Otub1 increased AT8-positive Tau in vivo, simi- homogenates from TauP301S transgenic mice injected with larly as wild-type Otub1 (Fig. 8). AAV–Otub1 and control mice injected with AAV–GFP were To further analyze the requirement for catalytically analyzed under nonreducing condition. This revealed a signif- active Otub1 for induction of oligomeric Tau forms, brain icant increase of monomeric Tau and oligomeric Tau in AAV– extracts of AAV-infected TauP301S mice were biochemi- Otub1-injected mice compared with AAV–GFP-injected mice cally analyzed 2 months postinjection, under nonreducing (Fig. 7c). Similarly as in primary neurons, Tauc (a homo- or condition. This revealed significantly increased oligomeric heterotypic Tau complex) was robustly induced in AAV– Tau forms, following expression of wild-type Otub1, but Otub1- but not AAV–GFP-injected mice, strongly confirm- not following expression of GFP or of the catalytically ing Otub1-induced modulation of Tau in brain (Fig. 7c). To inactive form of Otub1 (C91A) (Fig. 8c). This was further further confirm the increased concentration of oligomeric Tau confirmed by dot-blot analysis using T22 antibody recog- forms following Otub1 expression, dot-blot analysis using nizing oligomeric Tau forms (Fig. 8d). These data indicate oligomeric Tau-specific antibody T22 was performed. This that expression of catalytically active Otub1 is required for revealed significantly increased oligomeric Tau in AAV– accumulation of phosphorylated Tau and oligomeric Tau Otub1-injected mice compared with AAV–GFP-injected con- forms. trol cases (Fig. 7d). Noteworthy, to analyze whether endogenous Otub1 is modulated in conditions of accumulating pathological Tau The effect of Otub1 on Tau phosphorylation forms, we took advantage of our microarray analysis per- and oligomerization in vivo is dependent on its catalytic formed for a different project in which we analyzed time- function dependent changes in gene expression in hippocampus following Tau-seeded Tau aggregation in Tau transgenic mice. To further assess the requirement for catalytically active Interestingly, Otub1 expression is gradually downregulated Otub1 and the presence of the catalytically conserved C91 with increasing Tau-seeded Tau aggregation (Fig. S4a), which of Otub1, for accumulation of pathological Tau forms could be considered as a protective mechanism. Furthermore,

1 3 Acta Neuropathol

1 3 Acta Neuropathol analysis of hippocampal gene expression in aging mice, using Fig. 8 The effect of Otub1 on Tau phosphorylation and oligomeri-▸ publicly available data (NCBI database GDS2082) [73], indi- zation in vivo is dependent on its catalytic function. a Immunohis- tochemical staining of cortex of TauP301S transgenic mice injected cates a significant increase in Otub1 expression with aging with AAV–Otub1 (WT), catalytically dead mutant (C91A) or (Fig. S4b). The latter suggests that increased expression of N-terminal truncation (N-T) at P0 and analyzed at 2 months. Rep- Otub1 with aging might contribute to increased risk for accu- resentative images of AAV-driven expression of Otub1 (wild type, mulation of pathological Tau forms with aging. C91A, NT) are presented (green). Representative AT8 stainings are presented (red) for each group (3–5 sections were examined for each mouse; 9–12 fields were quantified for each section; scale bar 400 μm, ***p value <0.001). b Western blotting analysis was Discussion used to analyze phosphorylated Tau in cortex homogenates of mice injected with AAV–Otub1 wild type or two mutants, AAV–Otub1 C91A and AAV–Otub1 N-T, using AT8 (Ser202/Thr205) antibody Accumulation of aberrantly folded proteins is common to (n 3 mice for each group; ***p value <0.001). c Western blotting many neurodegenerative disorders, including AD, mak- analysis= of cortex homogenates from TauP301S mice injected with ing (dys)regulation of proteostasis not only a potentially AAV–Otub1 wild type, with AAV–Otub1 C91A, or AAV–GFP in implicated pathogenetic mechanism, but most impor- nonreducing condition using Tau antibody is presented. This reveals a clear increase in oligomeric forms of Tau only in TauP301S mice tantly an attractive therapeutic target. In AD, symptom injected with AAV–Otub1 wild type, while not in AAV–Otub1 C91A- progression strongly correlates with progression of Tau or AAV–GFP-injected TauP301S mice (n 3 mice for each group; pathology, with early, soluble, oligomeric forms of Tau *p value <0.05; **p value <0.01). Monomeric= Tau increased, less generally considered as toxic culprits. Aberrant protein strongly compared with oligomeric Tau, in TauP301S mice injected with AAV–Otub1 wild type, compared with AAV–Otub1 C91A- or degradation occurs through either the ALS or UPS, with AAV–GFP-injected TauP301S mice (n 3 mice for each group; the former mainly involved in degradation of large insol- *p value <0.05; ***p value <0.001). d =Dot-blot analysis of cortex uble protein aggregates, while the latter is preferentially homogenates from TauP301S mice injected with AAV–Otub1 wild involved in clearance of smaller soluble forms, being the type, with AAV–Otub1 C91A, or AAV–GFP using oligomeric Tau- specific antibody T22 is presented. Quantitative analysis was per- focus of this work. In view of the importance of the UPS formed following normalization to total Tau (dot-blot analysis) (n 3 in AD, Tau-ubiquitinating enzymes have been identified, mice for each group; **p value <0.01) = while Tau deubiquitination has received less attention. Starting from a Tau interactome mapping, we identified here for the first time Otub1 as a novel Tau deubiquit- and immunological analysis [13, 49–51, 56]. Interestingly, inase, affecting accumulation of pathological forms of we here identified Otub1 as a Tau deubiquitinase which Tau in vitro and in vivo. decreases the levels of Lys48- but not Lys63-linked poly- Here, we present a proteome-wide screening approach ubiquitin chains on Tau, increasing Tau stability in neurons. to identify Tau-interacting proteins, as a basis to gain Our finding is in line with previous reports showing that insight into Tau pathophysiology and to identify Tau modi- Otub1 has preference for Lys48-linked polyubiquitination fiers with therapeutic potential. Proteome-wide screening [17, 48, 74] and functions as a regulator of protein turnover approaches are increasingly used—in analogy to genome- [23, 26, 32, 44, 66]. Mutation of the catalytically conserved wide screening approaches—to gain novel insights from an cysteine (C91) abolished its role in Tau polyubiquitination, unbiased starting point for drug discovery and fundamen- further confirming the crucial role of its catalytic capacity tal science [3, 10, 27, 58, 62]. We used an iTRAQ-based for its involvement in the regulation of Lys48 polyubiquit- approach to identify proteins interacting with endogenous ination on Tau. Otub1 has also been found to have a nonca- Tau from mouse brain. The validity of our approach is nonical role in DNA damage response by impeding Lys63 reflected in the identification of several well-characterized polyubiquitin chain formation [53, 77, 78], but no effect of Tau-interacting proteins and the current identification of Otub1 on Lys63 polyubiquitination of Tau nor implication Otub1 as a Tau deubiquitinase. The presented Tau interac- of the N-terminal part of Otub1 was demonstrated. tome provides important information, extending beyond the Notably, the effect of polyubiquitination modification is current work, providing the basis for novel insights into the determinant for the fate of its substrate. Lys48 polyubiq- pathological role of Tau and to identify novel Tau-directed uitin chains act as a proteasomal degradative signal, while targets. Lys63-linked ubiquitin modification has been proposed Here, we validated this approach and identified for the to contribute to biogenesis of inclusions and clearance of first time Otub1 as a novel Tau deubiquitinase, based on the larger aggregates by autophagy [36, 55, 68, 80]. Although Tau interactome map. Accumulating evidence implicates soluble Tau forms as well as large insoluble aggregated Tau dysregulated proteostasis and a dysregulated UPS in AD, as forms can be polyubiquitinated [13, 49, 50], higher-order highlighted in the “Introduction.” Different types of poly- protein aggregations are unlikely to pass through the nar- ubiquitination of Tau, including Lys48 and Lys63 linked, row proteasome opening, and they are normally seques- have been identified in AD patients by mass spectrometry trated into inclusions or aggresomes and cleared through

1 3 Acta Neuropathol

the ALS. Recently, several lines of evidence have indi- In our experiments, we found that Otub1 only influences cated selective autophagy using Lys63-linked polyubiq- Lys48 polyubiquitination but not Lys63-linked polyubiqui- uitin chains as a cargo recognition signal [36, 55, 68, 80]. tin chains of Tau, and therefore would be rather involved

1 3 Acta Neuropathol in clearance of smaller soluble forms of Tau through the target. It must be noted that upstream modulators of Otub1, proteasome. mechanisms regulating Otub1 activity, expression and These findings are further strengthened by our assess- probably the binding of Otub1 to chaperone molecules ment of the effect of Otub1 on Tau metabolism. We demon- must be considered as potential therapeutic targets as well. strate that Otub1-dependent Tau deubiquitination is linked Regulation of key proteins via recruitment of distinct E3 to accumulation of Tau phosphorylated at the pathologi- ligases or DUBs mediated by specific adaptor proteins is cally relevant AT8 epitope—used for Braak staging—and common in many cellular pathways. While the mechanisms accumulation of oligomeric soluble forms of Tau, which upstream of Otub1 remain unknown, they may provide have been considered to be toxic culprits. Hence, overex- additional therapeutic targets. Inhibition of deubiquitinases pressed Otub1, which regulates Lys48-linked polyubiq- as therapeutic targets is attracting increasing attention for uitination of Tau, promotes accumulation of abnormally different diseases, including cancer. Our findings identify- phosphorylated Tau and oligomeric Tau, both in Tau trans- ing Otub1 as a Tau deubiquitinase affecting pathological genic mice and in primary neurons. These findings are in forms of Tau in vitro and in vivo may therefore yield new line with previously published findings. Depletion of CHIP perspectives for therapy. in Tau mice leads to significant accumulation of ubiquitin- Taken together, in this work we present a Tau interac- negative, and phosphorylated soluble Tau species [16]. In tome map, which yields a basis for novel insights into Tau a nonneuronal cell line which overexpresses the longest biology and its pathological role, validated by and extend- human Tau isoform, proteasome inhibitor treatment sta- ing beyond our current work. We furthermore provide novel bilizes phosphorylated and aggregated Tau species which insight into the link between the UPS and AD and open arise from Tau phosphatase PPA inactivation and normally new avenues for development of small-molecule inhibitors decay within 24 h [21]. In a triple transgenic mouse model, specifically targeting accumulation of pathological forms amyloid β-peptide (Aβ) immunotherapy leads to clearance of Tau in AD and related Tauopathies. of early soluble Tau forms by the UPS, but not of late Tau aggregates [54]. Along the same vein, molecular chaperon Acknowledgements This work was supported by the Belgian Fonds FKBP51, together with Hsp90, increase Tau oligomer for- National pour la Recherche Scientifique–Fonds de la Recherche Sci- entifique (FNRS-FRS; Qualified Researcher, Impulse Financing, mation by inhibiting Tau degradation through the protea- Research Credits), by Interuniversity Attraction Poles Programme– some in a mouse model [8]. In line with these reports, our Belgian State–Belgian Science Policy, by the Belgian Fonds de la work corroborates regulation of the early pathological Tau Recherche Scientifique Médicale, by the Queen Elisabeth Medical forms by the UPS, including deubiquitination as an impor- Foundation of Belgium (QEMF-FMRE), by the Stichting Alzheimer Onderzoek (SAO), by the Institute for the Promotion of Innovation tant modulator. by Science and Technology (IWT) in Flanders (IWT O&O, currently Our data are important in view of the fact that early oli- FWO), Belgium, and by Stellar funding of Janssen Research Founda- gomeric forms of Tau are generally considered to be toxic tion. We would like to thank Markus Boesche, Carola Doce, Frank culprits in AD and related Tauopathies. We have previ- Fischer, and Melanie Jundt for their technical expertise. ously shown in a Tau seeding assay in in vitro and in vivo Open Access This article is distributed under the terms of the Crea- models that neuronal dysfunction correlated strongly with tive Commons Attribution 4.0 International License (http://crea- early pathological forms of Tau rather than with full- tivecommons.org/licenses/by/4.0/), which permits unrestricted use, blown mature NFTs [65]. Previous studies using elegant distribution, and reproduction in any medium, provided you give approaches have indicated that the Tau aggregation process appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were and early pathological forms rather than full-blown NFTs made. correlate with neuronal dysfunction [6, 11, 12, 15, 33, 37, 39, 40, 59, 60, 67]. The detrimental role of Tau oligomers in neuronal function and behavior was further demonstrated References using acute injection of oligomeric Tau and immunization with anti-Tau oligomeric antibodies [11, 12, 39, 40]. Taken 1. Babu JR, Geetha T, Wooten MW (2005) Sequestosome 1/p62 together, these data highlight the importance of targeting shuttles polyubiquitinated tau for proteasomal degradation. J early oligomeric Tau forms as pathogenetic culprits, and Neurochem 94:192–203. doi:10.1111/j.1471-4159.2005.03181.x 2. Balakirev MY, Tcherniuk SO, Jaquinod M, Chroboczek J (2003) hence the importance of the demonstrated effect of Otub1 Otubains: a new family of cysteine proteases in the ubiqui- on soluble oligomeric forms of Tau. tin pathway. EMBO Rep 4:517–522. doi:10.1038/sj.embor. Interestingly, we show that the conserved catalytic embor824 cysteine (C91) of Otub1 is crucial for accumulation of 3. Bantscheff M, Eberhard D, Abraham Y, Bastuck S, Boesche M, Hobson S, Mathieson T, Perrin J, Raida M, Rau C, Reader hyperphosphorylated Tau and oligomeric Tau species. This V, Sweetman G, Bauer A, Bouwmeester T, Hopf C, Kruse U, finding is not only important as an internal control but also Neubauer G, Ramsden N, Rick J, Kuster B, Drewes G (2007) in the context of drug development, providing a druggable Quantitative chemical proteomics reveals mechanisms of action

1 3 Acta Neuropathol

of clinical ABL kinase inhibitors. Nat Biotechnol 25:1035–1044. 17. Edelmann MJ, Iphofer A, Akutsu M, Altun M, di Gleria K, doi:10.1038/nbt1328 Kramer HB, Fiebiger E, Dhe-Paganon S, Kessler BM (2009) 4. Bantscheff M, Hopf C, Savitski MM, Dittmann A, Grandi P, Structural basis and specificity of human otubain 1-medi- Michon AM, Schlegl J, Abraham Y, Becher I, Bergamini G, Boe- ated deubiquitination. Biochem J 418:379–390. doi:10.1042/ sche M, Delling M, Dumpelfeld B, Eberhard D, Huthmacher BJ20081318 C, Mathieson T, Poeckel D, Reader V, Strunk K, Sweetman G, 18. Eroglu B, Moskophidis D, Mivechi NF (2010) Loss of Hsp110 Kruse U, Neubauer G, Ramsden NG, Drewes G (2011) Chem- leads to age-dependent tau hyperphosphorylation and early accu- oproteomics profiling of HDAC inhibitors reveals selective mulation of insoluble amyloid beta. Mol Cell Biol 30:4626– targeting of HDAC complexes. Nat Biotechnol 29:255–265. 4643. doi:10.1128/MCB.01493-09 doi:10.1038/nbt.1759 19. Flach K, Ramminger E, Hilbrich I, Arsalan-Werner A, Albrecht 5. Bantscheff M, Schirle M, Sweetman G, Rick J, Kuster B F, Herrmann L, Goedert M, Arendt T, Holzer M (2014) Axotro- (2007) Quantitative mass spectrometry in proteomics: a criti- phin/MARCH7 acts as an E3 ubiquitin ligase and ubiquitinates cal review. Anal Bioanal Chem 389:1017–1031. doi:10.1007/ tau protein in vitro impairing microtubule binding. Biochim Bio- s00216-007-1486-6 phys Acta 1842:1527–1538. doi:10.1016/j.bbadis.2014.05.029 6. Berger Z, Roder H, Hanna A, Carlson A, Rangachari V, Yue M, 20. Giannakopoulos P, Herrmann FR, Bussiere T, Bouras C, Kovari Wszolek Z, Ashe K, Knight J, Dickson D, Andorfer C, Rosen- E, Perl DP, Morrison JH, Gold G, Hof PR (2003) Tangle and berry TL, Lewis J, Hutton M, Janus C (2007) Accumulation neuron numbers, but not amyloid load, predict cognitive status in of pathological tau species and memory loss in a conditional Alzheimer’s disease. Neurology 60:1495–1500 model of tauopathy. J Neurosci 27:3650–3662. doi:10.1523/ 21. Goldbaum O, Oppermann M, Handschuh M, Dabir D, Zhang JNEUROSCI.0587-07.2007 B, Forman MS, Trojanowski JQ, Lee VM, Richter-Landsberg C 7. Berke SJ, Paulson HL (2003) Protein aggregation and the ubiqui- (2003) Proteasome inhibition stabilizes tau inclusions in oligo- tin proteasome pathway: gaining the UPPer hand on neurodegen- dendroglial cells that occur after treatment with okadaic acid. J eration. Curr Opin Genet Dev 13:253–261 Neurosci 23:8872–8880 8. Blair LJ, Nordhues BA, Hill SE, Scaglione KM, O’Leary JC 3rd, 22. Goldberg AL (2003) Protein degradation and protection Fontaine SN, Breydo L, Zhang B, Li P, Wang L, Cotman C, Paul- against misfolded or damaged proteins. Nature 426:895–899. son HL, Muschol M, Uversky VN, Klengel T, Binder EB, Kayed doi:10.1038/nature02263 R, Golde TE, Berchtold N, Dickey CA (2013) Accelerated neu- 23. Goncharov T, Niessen K, de Almagro MC, Izrael-Tomasevic A, rodegeneration through chaperone-mediated oligomerization of Fedorova AV, Varfolomeev E, Arnott D, Deshayes K, Kirkpatrick tau. J Clin Invest 123:4158–4169. doi:10.1172/JCI69003 DS, Vucic D (2013) OTUB1 modulates c-IAP1 stability to regu- 9. Braak H, Braak E (1991) Neuropathological stageing of Alzhei- late signalling pathways. EMBO J 32:1103–1114. doi:10.1038/ mer-related changes. Acta Neuropathol 82:239–259 emboj.2013.62 10. Brajenovic M, Joberty G, Kuster B, Bouwmeester T, Drewes G 24. Guo JL, Lee VM (2011) Seeding of normal Tau by pathological (2004) Comprehensive proteomic analysis of human Par protein Tau conformers drives pathogenesis of Alzheimer-like tangles. J complexes reveals an interconnected protein network. J Biol Biol Chem 286:15317–15331. doi:10.1074/jbc.M110.209296 Chem 279:12804–12811. doi:10.1074/jbc.M312171200 25. Han DH, Na HK, Choi WH, Lee JH, Kim YK, Won C, Lee SH, 11. Castillo-Carranza DL, Gerson JE, Sengupta U, Guerrero-Munoz Kim KP, Kuret J, Min DH, Lee MJ (2014) Direct cellular deliv- MJ, Lasagna-Reeves CA, Kayed R (2014) Specific targeting of ery of human proteasomes to delay tau aggregation. Nat Com- tau oligomers in Htau mice prevents cognitive impairment and mun 5:5633. doi:10.1038/ncomms6633 tau toxicity following injection with brain-derived tau oligomeric 26. Herhaus L, Al-Salihi M, Macartney T, Weidlich S, Sapkota GP seeds. J Alzheimers Dis 40(Suppl 1):S97–S111. doi:10.3233/ (2013) OTUB1 enhances TGFbeta signalling by inhibiting the JAD-132477 ubiquitylation and degradation of active SMAD2/3. Nat Com- 12. Castillo-Carranza DL, Sengupta U, Guerrero-Munoz MJ, Lasa- mun 4:2519. doi:10.1038/ncomms3519 gna-Reeves CA, Gerson JE, Singh G, Estes DM, Barrett AD, 27. Huber KV, Olek KM, Muller AC, Tan CS, Bennett KL, Colinge Dineley KT, Jackson GR, Kayed R (2014) Passive immuniza- J, Superti-Furga G (2015) Proteome-wide drug and metabolite tion with Tau oligomer monoclonal antibody reverses tauopa- interaction mapping by thermal-stability profiling. Nat Methods thy phenotypes without affecting hyperphosphorylated neu- 12:1055–1057. doi:10.1038/nmeth.3590 rofibrillary tangles. J Neurosci 34:4260–4272. doi:10.1523/ 28. Ihara Y, Morishima-Kawashima M, Nixon R (2012) The ubiq- JNEUROSCI.3192-13.2014 uitin-proteasome system and the autophagic-lysosomal sys- 13. Cripps D, Thomas SN, Jeng Y, Yang F, Davies P, Yang AJ (2006) tem in Alzheimer disease. Cold Spring Harb Perspect Med 2. Alzheimer disease-specific conformation of hyperphosphoryl- doi:10.1101/cshperspect.a006361 ated paired helical filament-Tau is polyubiquitinated through 29. Irmler M, Gentier RJ, Dennissen FJ, Schulz H, Bolle I, Holter Lys-48, Lys-11, and Lys-6 ubiquitin conjugation. J Biol Chem SM, Kallnik M, Cheng JJ, Klingenspor M, Rozman J, Ehrhardt 281:10825–10838. doi:10.1074/jbc.M512786200 N, Hermes DJ, Gailus-Durner V, Fuchs H, Hrabe de Ange- 14. David DC, Layfield R, Serpell L, Narain Y, Goedert M, Spillan- lis M, Meyer HE, Hopkins DA, Van Leeuwen FW, Beckers J tini MG (2002) Proteasomal degradation of tau protein. J Neuro- (2012) Long-term proteasomal inhibition in transgenic mice by chem 83:176–185 UBB( 1) expression results in dysfunction of central respira- 15. Deger JM, Gerson JE, Kayed R (2015) The interrelationship of tion control+ reminiscent of brainstem neuropathology in Alz- proteasome impairment and oligomeric intermediates in neuro- heimer patients. Acta Neuropathol 124:187–197. doi:10.1007/ degeneration. Aging Cell 14:715–724. doi:10.1111/acel.12359 s00401-012-1003-7 16. Dickey CA, Yue M, Lin WL, Dickson DW, Dunmore JH, Lee 30. Jinwal UK, O’Leary JC 3rd, Borysov SI, Jones JR, Li Q, Koren WC, Zehr C, West G, Cao S, Clark AM, Caldwell GA, Cald- J 3rd, Abisambra JF, Vestal GD, Lawson LY, Johnson AG, Blair well KA, Eckman C, Patterson C, Hutton M, Petrucelli L (2006) LJ, Jin Y, Miyata Y, Gestwicki JE, Dickey CA (2010) Hsc70 rap- Deletion of the ubiquitin ligase CHIP leads to the accumula- idly engages tau after microtubule destabilization. J Biol Chem tion, but not the aggregation, of both endogenous phospho- 285:16798–16805. doi:10.1074/jbc.M110.113753 and caspase-3-cleaved tau species. J Neurosci 26:6985–6996. 31. Juang YC, Landry MC, Sanches M, Vittal V, Leung CC, Cec- doi:10.1523/JNEUROSCI.0746-06.2006 carelli DF, Mateo AR, Pruneda JN, Mao DY, Szilard RK,

1 3 Acta Neuropathol

Orlicky S, Munro M, Brzovic PS, Klevit RE, Sicheri F, Duro- 48. Messick TE, Russell NS, Iwata AJ, Sarachan KL, Shiekhattar cher D (2012) OTUB1 co-opts Lys48-linked ubiquitin recogni- R, Shanks JR, Reyes-Turcu FE, Wilkinson KD, Marmorstein tion to suppress E2 enzyme function. Mol Cell 45:384–397. R (2008) Structural basis for ubiquitin recognition by the Otu1 doi:10.1016/j.molcel.2012.01.011 ovarian tumor domain protein. J Biol Chem 283:11038–11049. 32. Karunarathna U, Kongsema M, Zona S, Gong C, Cabrera E, doi:10.1074/jbc.M704398200 Gomes AR, Man EP, Khongkow P, Tsang JW, Khoo US, Medema 49. Mori H, Kondo J, Ihara Y (1987) Ubiquitin is a component RH, Freire R, Lam EW (2016) OTUB1 inhibits the ubiquitination of paired helical filaments in Alzheimer’s disease. Science and degradation of FOXM1 in breast cancer and epirubicin resist- 235:1641–1644 ance. Oncogene 35:1433–1444. doi:10.1038/onc.2015.208 50. Morishima-Kawashima M, Hasegawa M, Takio K, Suzuki M, 33. Kayed R, Jackson GR (2009) Prefilament tau species as poten- Titani K, Ihara Y (1993) Ubiquitin is conjugated with amino- tial targets for immunotherapy for Alzheimer disease and related terminally processed tau in paired helical filaments. Neuron disorders. Curr Opin Immunol 21:359–363. doi:10.1016/j. 10:1151–1160 coi.2009.05.001 51. Morris M, Knudsen GM, Maeda S, Trinidad JC, Ioanoviciu A, 34. Keck S, Nitsch R, Grune T, Ullrich O (2003) Proteasome inhibi- Burlingame AL, Mucke L (2015) Tau post-translational modifi- tion by paired helical filament-tau in brains of patients with Alz- cations in wild-type and human amyloid precursor protein trans- heimer’s disease. J Neurochem 85:115–122 genic mice. Nat Neurosci 18:1183–1189. doi:10.1038/nn.4067 35. Keller JN, Gee J, Ding Q (2002) The proteasome in brain aging. 52. Myeku N, Clelland CL, Emrani S, Kukushkin NV, Yu WH, Ageing Res Rev 1:279–293 Goldberg AL, Duff KE (2016) Tau-driven 26S proteasome 36. Kraft C, Peter M, Hofmann K (2010) Selective autophagy: ubiq- impairment and cognitive dysfunction can be prevented early in uitin-mediated recognition and beyond. Nat Cell Biol 12:836– disease by activating cAMP-PKA signaling. Nat Med 22:46–53. 841. doi:10.1038/ncb0910-836 doi:10.1038/nm.4011 37. Kruger L, Mandelkow EM (2016) Tau neurotoxicity and rescue 53. Nakada S, Tai I, Panier S, Al-Hakim A, Iemura S, Juang YC, in animal models of human Tauopathies. Curr Opin Neurobiol O’Donnell L, Kumakubo A, Munro M, Sicheri F, Gingras AC, 36:52–58. doi:10.1016/j.conb.2015.09.004 Natsume T, Suda T, Durocher D (2010) Non-canonical inhibition 38. Lam YA, Pickart CM, Alban A, Landon M, Jamieson C, Ramage of DNA damage-dependent ubiquitination by OTUB1. Nature R, Mayer RJ, Layfield R (2000) Inhibition of the ubiquitin-pro- 466:941–946. doi:10.1038/nature09297 teasome system in Alzheimer’s disease. Proc Natl Acad Sci USA 54. Oddo S, Billings L, Kesslak JP, Cribbs DH, LaFerla FM (2004) 97:9902–9906. doi:10.1073/pnas.170173897 Abeta immunotherapy leads to clearance of early, but not late, 39. Lasagna-Reeves CA, Castillo-Carranza DL, Guerrero-Muoz MJ, hyperphosphorylated tau aggregates via the proteasome. Neuron Jackson GR, Kayed R (2010) Preparation and characterization 43:321–332. doi:10.1016/j.neuron.2004.07.003 of neurotoxic tau oligomers. Biochemistry 49:10039–10041. 55. Olzmann JA, Chin LS (2008) Parkin-mediated K63-linked poly- doi:10.1021/bi1016233 ubiquitination: a signal for targeting misfolded proteins to the 40. Lasagna-Reeves CA, Castillo-Carranza DL, Sengupta U, Sarm- aggresome-autophagy pathway. Autophagy 4:85–87 iento J, Troncoso J, Jackson GR, Kayed R (2012) Identification 56. Paine S, Bedford L, Thorpe JR, Mayer RJ, Cavey JR, Bajaj N, of oligomers at early stages of tau aggregation in Alzheimer’s Sheppard PW, Lowe J, Layfield R (2009) Immunoreactivity to disease. FASEB J 26:1946–1959. doi:10.1096/fj.11-199851 Lys63-linked polyubiquitin is a feature of neurodegeneration. 41. Lee G, Neve RL, Kosik KS (1989) The microtubule binding Neurosci Lett 460:205–208. doi:10.1016/j.neulet.2009.05.074 domain of tau protein. Neuron 2:1615–1624 57. Petrucelli L, Dickson D, Kehoe K, Taylor J, Snyder H, Grover A, 42. Leroy E, Boyer R, Auburger G, Leube B, Ulm G, Mezey E, De Lucia M, McGowan E, Lewis J, Prihar G, Kim J, Dillmann Harta G, Brownstein MJ, Jonnalagada S, Chernova T, Dehejia A, WH, Browne SE, Hall A, Voellmy R, Tsuboi Y, Dawson TM, Lavedan C, Gasser T, Steinbach PJ, Wilkinson KD, Polymero- Wolozin B, Hardy J, Hutton M (2004) CHIP and Hsp70 regulate poulos MH (1998) The ubiquitin pathway in Parkinson’s disease. tau ubiquitination, degradation and aggregation. Hum Mol Genet Nature 395:451–452. doi:10.1038/26652 13:703–714. doi:10.1093/hmg/ddh083 43. Levine B, Klionsky DJ (2004) Development by self-digestion: 58. Reinhard FB, Eberhard D, Werner T, Franken H, Childs D, Doce molecular mechanisms and biological functions of autophagy. C, Savitski MF, Huber W, Bantscheff M, Savitski MM, Drewes Dev Cell 6:463–477 G (2015) Thermal proteome profiling monitors ligand interac- 44. Li S, Zheng H, Mao AP, Zhong B, Li Y, Liu Y, Gao Y, Ran Y, tions with cellular membrane proteins. Nat Methods 12:1129– Tien P, Shu HB (2010) Regulation of virus-triggered signaling 1131. doi:10.1038/nmeth.3652 by OTUB1- and OTUB2-mediated deubiquitination of TRAF3 59. Roberson ED, Scearce-Levie K, Palop JJ, Yan F, Cheng IH, Wu and TRAF6. J Biol Chem 285:4291–4297. doi:10.1074/jbc. T, Gerstein H, Yu GQ, Mucke L (2007) Reducing endogenous M109.074971 tau ameliorates amyloid beta-induced deficits in an Alzhei- 45. Lindsten K, de Vrij FM, Verhoef LG, Fischer DF, van Leeuwen mer’s disease mouse model. Science 316:750–754. doi:10.1126/ FW, Hol EM, Masucci MG, Dantuma NP (2002) Mutant ubiq- science.1141736 uitin found in neurodegenerative disorders is a ubiquitin fusion 60. Santacruz K, Lewis J, Spires T, Paulson J, Kotilinek L, Ingelsson degradation substrate that blocks proteasomal degradation. J Cell M, Guimaraes A, DeTure M, Ramsden M, McGowan E, Forster Biol 157:417–427. doi:10.1083/jcb.200111034 C, Yue M, Orne J, Janus C, Mariash A, Kuskowski M, Hyman B, 46. Luo HB, Xia YY, Shu XJ, Liu ZC, Feng Y, Liu XH, Yu G, Yin Hutton M, Ashe KH (2005) Tau suppression in a neurodegenera- G, Xiong YS, Zeng K, Jiang J, Ye K, Wang XC, Wang JZ (2014) tive mouse model improves memory function. Science 309:476– SUMOylation at K340 inhibits tau degradation through deregu- 481. doi:10.1126/science.1113694 lating its phosphorylation and ubiquitination. Proc Natl Acad Sci 61. Sato Y, Yamagata A, Goto-Ito S, Kubota K, Miyamoto R, Nakada USA 111:16586–16591. doi:10.1073/pnas.1417548111 S, Fukai S (2012) Molecular basis of Lys-63-linked polyubiq- 47. Magnani E, Fan J, Gasparini L, Golding M, Williams M, Schiavo uitination inhibition by the interaction between human deubiq- G, Goedert M, Amos LA, Spillantini MG (2007) Interaction of uitinating enzyme OTUB1 and ubiquitin-conjugating enzyme tau protein with the dynactin complex. EMBO J 26:4546–4554. UBC13. J Biol Chem 287:25860–25868. doi:10.1074/jbc. doi:10.1038/sj.emboj.7601878 M112.364752

1 3 Acta Neuropathol

62. Savitski MM, Reinhard FB, Franken H, Werner T, Savitski MF, 71. van Leeuwen FW, de Kleijn DP, van den Hurk HH, Neubauer A, Eberhard D, Martinez Molina D, Jafari R, Dovega RB, Klaeger Sonnemans MA, Sluijs JA, Koycu S, Ramdjielal RD, Salehi A, S, Kuster B, Nordlund P, Bantscheff M, Drewes G (2014) Track- Martens GJ, Grosveld FG, Peter J, Burbach H, Hol EM (1998) ing cancer drugs in living cells by thermal profiling of the pro- Frameshift mutants of beta amyloid precursor protein and ubiq- teome. Science 346:1255784. doi:10.1126/science.1255784 uitin-B in Alzheimer’s and Down patients. Science 279:242–247 63. Shimura H, Schwartz D, Gygi SP, Kosik KS (2004) CHIP- 72. Vasconcelos B, Stancu IC, Buist A, Bird M, Wang P, Vanoosthuyse Hsc70 complex ubiquitinates phosphorylated tau and enhances A, Van Kolen K, Verheyen A, Kienlen-Campard P, Octave JN, cell survival. J Biol Chem 279:4869–4876. doi:10.1074/jbc. Baatsen P, Moechars D, Dewachter I (2016) Heterotypic seeding M305838200 of Tau fibrillization by pre-aggregated Abeta provides potent seeds 64. Stancu IC, Ris L, Vasconcelos B, Marinangeli C, Goeminne for prion-like seeding and propagation of Tau-pathology in vivo. L, Laporte V, Haylani LE, Couturier J, Schakman O, Gailly P, Acta Neuropathol 131:549–569. doi:10.1007/s00401-015-1525-x Pierrot N, Kienlen-Campard P, Octave JN, Dewachter I (2014) 73. Verbitsky M, Yonan AL, Malleret G, Kandel ER, Gilliam TC, Tauopathy contributes to synaptic and cognitive deficits in a Pavlidis P (2004) Altered hippocampal transcript profile accom- murine model for Alzheimer’s disease. FASEB J 28:2620–2631. panies an age-related spatial memory deficit in mice. Learn Mem doi:10.1096/fj.13-246702 11:253–260. doi:10.1101/lm.68204 65. Stancu IC, Vasconcelos B, Ris L, Wang P, Villers A, Peeraer 74. Wang T, Yin L, Cooper EM, Lai MY, Dickey S, Pickart CM, E, Buist A, Terwel D, Baatsen P, Oyelami T, Pierrot N, Cas- Fushman D, Wilkinson KD, Cohen RE, Wolberger C (2009) teels C, Bormans G, Kienlen-Campard P, Octave JN, Moe- Evidence for bidentate substrate binding as the basis for the K48 chars D, Dewachter I (2015) Templated misfolding of Tau by linkage specificity of otubain 1. J Mol Biol 386:1011–1023. prion-like seeding along neuronal connections impairs neuronal doi:10.1016/j.jmb.2008.12.085 network function and associated behavioral outcomes in Tau 75. Weingarten MD, Lockwood AH, Hwo SY, Kirschner MW transgenic mice. Acta Neuropathol 129:875–894. doi:10.1007/ (1975) A protein factor essential for microtubule assembly. Proc s00401-015-1413-4 Natl Acad Sci USA 72:1858–1862 66. Sun XX, Challagundla KB, Dai MS (2012) Positive regulation 76. Werner T, Becher I, Sweetman G, Doce C, Savitski MM, of p53 stability and activity by the deubiquitinating enzyme Otu- Bantscheff M (2012) High-resolution enabled TMT 8-plexing. bain 1. EMBO J 31:576–592. doi:10.1038/emboj.2011.434 Anal Chem 84:7188–7194. doi:10.1021/ac301553x 67. Sydow A, Van der Jeugd A, Zheng F, Ahmed T, Balschun D, 77. Wiener R, DiBello AT, Lombardi PM, Guzzo CM, Zhang X, Petrova O, Drexler D, Zhou L, Rune G, Mandelkow E, D’Hooge Matunis MJ, Wolberger C (2013) E2 ubiquitin-conjugating R, Alzheimer C, Mandelkow EM (2011) Tau-induced defects in enzymes regulate the deubiquitinating activity of OTUB1. Nat synaptic plasticity, learning, and memory are reversible in trans- Struct Mol Biol 20:1033–1039. doi:10.1038/nsmb.2655 genic mice after switching off the toxic Tau mutant. J Neurosci 78. Wiener R, Zhang X, Wang T, Wolberger C (2012) The mecha- 31:2511–2525. doi:10.1523/JNEUROSCI.5245-10.2011 nism of OTUB1-mediated inhibition of ubiquitination. Nature 68. Tan JM, Wong ES, Kirkpatrick DS, Pletnikova O, Ko HS, Tay 483:618–622. doi:10.1038/nature10911 SP, Ho MW, Troncoso J, Gygi SP, Lee MK, Dawson VL, Daw- 79. Witman GB, Cleveland DW, Weingarten MD, Kirschner MW son TM, Lim KL (2008) Lysine 63-linked ubiquitination pro- (1976) Tubulin requires tau for growth onto microtubule initiat- motes the formation and autophagic clearance of protein inclu- ing sites. Proc Natl Acad Sci USA 73:4070–4074 sions associated with neurodegenerative diseases. Hum Mol 80. Wooten MW, Geetha T, Babu JR, Seibenhener ML, Peng J, Cox Genet 17:431–439. doi:10.1093/hmg/ddm320 N, Diaz-Meco MT, Moscat J (2008) Essential role of seques- 69. Thal DR, Rub U, Orantes M, Braak H (2002) Phases of A beta- tosome 1/p62 in regulating accumulation of Lys63-ubiquit- deposition in the human brain and its relevance for the develop- inated proteins. J Biol Chem 283:6783–6789. doi:10.1074/jbc. ment of AD. Neurology 58:1791–1800 M709496200 70. Van der Jeugd A, Hochgrafe K, Ahmed T, Decker JM, Sydow A, 81. Yoshiyama Y, Higuchi M, Zhang B, Huang SM, Iwata N, Saido Hofmann A, Wu D, Messing L, Balschun D, D’Hooge R, Man- TC, Maeda J, Suhara T, Trojanowski JQ, Lee VM (2007) Syn- delkow EM (2012) Cognitive defects are reversible in inducible apse loss and microglial activation precede tangles in a P301S mice expressing pro-aggregant full-length human Tau. Acta Neu- tauopathy mouse model. Neuron 53:337–351. doi:10.1016/j. ropathol 123:787–805. doi:10.1007/s00401-012-0987-3 neuron.2007.01.010

1 3